Home

Welcome to the Sequential_Fish package wiki!

Sequential Fish is an imaging technology relying on microfluidics systems to perform several rounds of FISH sequentialy. This technique allows visualisation of multiple RNAs in same cells yielding both spatial and quantitative informations as well as the possibility to correlate those informations betweens different single molecules distributions.
However, it also raises the critical need for a fully automated quantification pipeline and a set of analysis techniques. The Sequential_Fish package was built to answer this need and propose a few modules for seqFish data quantification.

Pipeline

• 2D/3D nuclei and cytoplasm segmentation using cellpose 4.+.
• Cycle alignement using Fourrier analysis on dapi signal.
• Chromatic abberations correction using polynomial interpolation.
• LoG Filter + Threshold 3D single molecule detection.
• Spot clustering algorithm (DBscan) for foci and transcription site detection.
• Bright region deconvolution to estimate single molecule number in foci/Tx.
• Indivual cell and individual spots quantification.

Viewer

The package comes in with a custom napari viewer allowing to browse through your data and to display your FISH signal as well as every results from the pipeline.

Analysis

All the raw data from quantification is available for custom analysis but you can also try out our analysis pipelines that are ready out of the box.

• Quality assessment / sum up of experiment and quantification
• Pairwise co-localization analysis
• Nco-localization analysis
• Quantitative distributions
• Single molecule expression correlation analysis

Requirements

Here are the requirements to run the Sequential_Fish package :

• Python3.12 installation
• ome.tiff file structure
• One channel contains dapi signal.
• 64GB + ram (depending on cycle numbers - this might evolve in the future as the pipeline is still in early stage of development and memory usage is not optimized)
• GPU set up with cellpose [Not compulsory but higly recommanded]